A folded schedule
Fits in your pocket; click to download PDF

Sound Transit’s June 2011 service change took effect today. As usual, Link has the basic span and frequency of service information in a schedule. For those who want more detail, I created this unofficial schedule [UPDATED: October 2013] that shows, to the minute, when frequency changes and the first and last trains of the day. It includes a fare table and line map with station-to-station travel times.

The reason I didn’t include times for every station in my schedule is because the pattern is very consistent and the math is simple. In fact, I’d argue that my schedule is unnecessary once you learn the pattern. All you need to know are the first and last train times from Westlake and SeaTac/Airport, the frequency schedule, and the travel time between stations. Knowing that, you can derive most of the detailed schedule from Sound Transit’s basic schedule.

As I spent time creating this schedule I saw why Sound Transit made its schedule the way it is. Of course, many people won’t spend time figuring this out, so a more detailed schedule than currently printed is still needed. All they need to do is add the timetables to the current format. I tried and it works, with some space left for other useful information like travel times and late night/early morning bus connections. Throw in some real-time arrival predictions at stations and the result is a significantly more user-friendly system than what we have now.

38 Replies to “An Unofficial Link Schedule”

  1. Looks good, but you should be using a 24 hour clock. I was confused for 2minutes until I saw the theme and then came across the note afterward.

    1. I was using it internally then switched to 12 hour clock because most people are not familiar with the 24 hour clock. Most people do know the convention that bold text equals PM as all the agencies around here do it on their timetables. For a while I had tiny AM PM symbols where it changes but removed them for lack of space.

      1. I suppose that convention just seems foreign to me. But I figured the rationale might be something like that. Perhaps we start the revolution here and force people to use a rational time system!

      2. Even changing to 24hr clocks won’t make our time system rational. How many kilo-seconds are in a day? How many mega-seconds in a year?

  2. One little suggestion. It looks like thematically you’re using Westlake, Stadium, Mt Baker, Ranier Beach and SeaTac as the primary stations. However, in the travel times graphic at the bottom you point to SODO instead of Stadium. It would be more consistent if you pointed to Stadium :)

  3. All my life, it’s been my dream to take a crap in a light rail restroom in Seattle with real-time arrival data in the cubicle. Sound Transit is destroying my dreams.

    1. Meantime, will settle for a pull-down seat on the toilet, like people usually use if they haven’t been sent to prison. I think it’s in the Bill of Rights somewhere. Or maybe that’s one of the powers reserved to states or individual citizens- in Washington State, hope it’s the second one.

      Will also settle for rest rooms. Might save on cleaning and sanitizing elevators.

      Meantime, having schedules is a good start. And while we’ve got VMS readouts, how about train arrival times there, too. BRT can do it. So see, you’ve got BRT and Oran both doing schedules. Tell us again: why, two years after opening, can’t Sound Transit do these things?

      Mark Dublin

  4. haha, great job on the graphics … looks exactly like the ferry schedules of my youf….

    ahhh natsukashiii

  5. Has Sound Transit ever given a credible response as to why they operate Link to a schedule, have a schedule for operators to follow, but won’t publish that schedule on the website or post it in the stations?

    1. I don’t know if they have ‘officially’ responded to that question, but it’s pretty obvious that there are two reasons.
      1. Timetable presentation is much longer and complex than just giving headway times.
      2. You can’t complain about a train being X minutes late at your stop if it’s not published. About the only thing you can say is “I waited so many minutes and the train should have shown up by then”. That solves a lot of problems.
      My observation is that Link doesn’t have the headways needed to afford that luxury of brevity. Train headways of less than 10 minutes only occur 30% of the day. The other 70% of the time it’s either 10 or 15 minute headways, which is longer than quite a few bus routes, or families of routes(e.g. 70 series), that DO publish a complete schedule.
      If ST had real time arrival info at the platform, I’d say it doesn’t really matter, but when your next train is 15 minutes away, ‘maybe’, and you don’t know if you just missed one, then it’s a real problem.
      Otis Elevator did a study one time on human behavior waiting for the next car. 2 minutes was barely noticed. At the 5 minute level people get anxious. At more than 5 minutes, the anxiety level rises pretty darn quick.
      Most people I know like the predictability of a schedule, and hoof it down to the stop a few minutes before the bus arrives.

      1. 1. There are multiple ways to present the information that minimizes complexity, and you can always provide both the abbreviated headways for those who don’t care for the complexity, and the schedule for those who do.

        2. That’s not a credible reason, that’s a desire to escape accountability.

        Real-time arrival information is useful for people who are already on the platform or are at a very short distance. It’s not useful for people who are trying to build a connection nor who are at some distance from the station. Like I need to leave the main terminal at Seatac about 7 minutes before the train is scheduled to leave to comfortably make it.

        By the way, what’s real-time information at a terminal station like Westlake or Seatac? It’s a schedule! Until a train starts its run there’s no real-time information, is there?

        Oran has a schedule. Link’s operators have a schedule. Sound Transit thumbs its nose at riders by not posting it.

      2. It might not be so bad if the published the headways AND had a clock in the station that didn’t say when the train was coming, but when the last one did!

      3. I completely agree with Carl. My own anxiety level starts at about the 5 minute mark, and by 10 minutes it’s a short fuse to the powder keg.
        Hey Ryan, if the clock just told me “Sucker!, you should have been her 2 minutes ago!”, I think I’d eventually throw rocks at the damn thing.

      4. I like the idea of the most-recent-arrival signs, such as for southbound MBS and northbound RBS. When it reaches eleven minutes, and no arrival announcement has been given, then you start to get a clue that Houston, we have a problem.

        Also, a next-train arrival sign wouldn’t do much for me. I’m already there, and won’t run off to the store just because I have seven minutes to kill. Now, a next-train arrival app …

    2. 1. The schedule is there. You just have to do some simple addition to get it.

      2. Sound Transit moved to a headway based on-time performance measurement. Therefore a train is now late if it deviates from the published headway regardless of some timepoint-based schedule. More on that and an official explanation in a future post.

      3. Transit industry standard is 10 minutes or less for random arrivals. Above that, people read schedules to reduce wait time. I never looked at a schedule for the 70 series buses before I leave but I do sometimes look at OneBusAway. 73% of surveyed OneBusAway users don’t read a published schedule anymore. Some even think the “schedule is a lie”. 7 minutes inside the airport is well within the useful reach of RT info. Portland and Vancouver, BC both have a highly visible from a far distance RT train info display inside their airports.

      4. How many people manually plan their trips anymore? Most of the time I just use Google Maps or Trip Planner. It’s far quicker and easier than pulling up multiple pages of schedules. If someone really wants to sit down and plan this out, they certainly have the time and mental capability to do the math (see 1).

      5. The wait time expectations of elevators are not comparable to transit service. For elevators, you expect get one more or less instantly after pushing that button. For transit service, that’s not the case. Real-time information reduces that anxiety.

      1. I still plan my trips manually. I’m a Luddite.

        The schedule is beautiful! I would have used Beacon Hill as the middle column, just to reduce confusion regarding the first/last trips.

        Way coolness would be listing the first and last transfers at non-downtown stations, if this were a larger unofficial pamphlet.

        I’d also add a way to send a donation to STB, but that’s just the political junkie in me.

      2. 1. Then publish it

        2. What does that mean at endpoints? If one leaves late, the next one can leave late? Outside of peak-period tunnel congestion or extraordinary incidents, I don’t see much variation, so I don’t think measuring headways vs. schedule make much difference, unless they are trying to paper over peak tunnel congestion – in which case I think they should expose it, not paper over it.

        3. I think I agree mainly that if it’s better than 10 minute headways, fine, but 10 minute headways are so easy to plan for, and so many connecting routes have 30 minute frequencies, so if I want to connect from Link to the 255 or 545 in Seattle in the evening, I want the Link schedule so that I can plan for a nice padded 5-10 minute connection, and not experience at 25 minute wait (or 55 minutes when they drop to hourly).

        4. Enough people plan that they shouldn’t be dismissed. And we vote and pay taxes – maybe more than non-planners.

      3. 1. See my link in the last paragraph. It’s feasible to do that in the printed schedules. However, I don’t think such detail is needed at the stations themselves. And ST takes a ridiculously long time to keep posted information at stations updated.

        2. By definition, there is no late as long as trains depart every 10 minutes or whatever the scheduled headway is. Anyway, waiting for explanation from ST on this one.

        3. You don’t need the Link schedule to plan for that though it is preferable to know to minimize wait time. I know the 255 departs IDS at 7:08 pm. I want a 5-minute pad. I know how long it takes between Airport and IDS: 32 minutes. Trains run every 10 minutes. So I arrive at the Airport station at 6:20 pm to guarantee a connection. If I knew the schedule, I’d arrive at 6:25 pm with time to spare for the 6:30 pm departure. I arrive downtown at 7:02 pm with 6 minutes of padding time. So I saved 5 minutes. Maybe that’s valuable. Maybe not. If I had RT info maybe I could shave a few minutes off.

        4. They do offer the Trip Planner and if you don’t have a computer you can always call customer service.

  6. Much as I like precision, since it runs so frequently, does a “schedule” really matter that much. What most people expect (I think) is to do things at their leisure, and when ready, go to the station and jump on train which should have a reasonably short maximum wait time…it ain’t the Orient Express you know.

    1. Yes it does. I don’t want a random wait between 0 & 15 minutes. I want to know if I should hurry to avoid missing a train, or have time for a coffee or restroom break. Maybe in the 7-8 minute range I don’t care and won’t hold them accountable. In the 10 & 15 minute headway periods, I want a schedule.

      1. That doesn’t apply if you want to make an appointment or a connection to a 30-minute bus.

      2. But you said “I want to know if I should hurry to avoid missing a train, or have time for a coffee or restroom break. ” That’s exactly what OBA is good for. You didn’t say I need to “make an appointment or a connection to a 30-minute bus.”

      3. When I’m leaving an airplane, I’d like to know the 10-minute clockface schedule, not have to get out my phone and have to try to connect to OBA, remember the stop #, etc. Just let me rely on the scheduled times.

      4. I think you should take the restroom break whether the train is close or not…you wouldn’t want to have an accident.

    2. What I’m seeing from these comments is that some people are positive that everyone, just like them, needs a schedule. And other people are positive that everyone, just like them, doesn’t.

      Unfortunately for us, we’re not all the same. :)

  7. Very clear, comprehensible and convenient, thanks. More information could be printed on the back of the page, too. I’d like to see ST take this start and print a schedule that shows an expanded weekday schedule on one side and the weekend information on the back side. I think it would be helpful to show at least one full hour of departure times during the rush hours or when service is every 15 minutes. Schedules like these probably work best when they presume that the reader is in the 5th grade.

    1. Another interesting feature would be a full station to station travel time matrix. This could be provided in a condensed form by combining with the fare table, with travel time in the upper right triangle and fares in the lower left triangle.

      1. I sort of thought of it in addition to the line map. I’m lazy when it comes to repeated addition. This is also why I’m okay with the way you’ve done the abbreviated schedule for ten and fifteen minute headways, but I’d prefer a full hour span of example times for the 7.5 minute headways. But as you said about the 10 minute headway rule, it doesn’t matter much for smaller headways.

      2. I’d love to but I’m keeping things simple. My printer doesn’t do automatic duplex and I always get frustrated when I try to print two sides on the same sheet, ending up printing it on the same side.

  8. Oran,

    I just want to say THANK YOU. I just moved into a house near the Mt. Baker station, and I was wondering if there was a way to get a schedule like this. As much as I love to do math, I’m glad you did it for me.


Comments are closed.